This paper carries out a research on AIGC personalized advertisement generation strategy and optimization of dissemination effect. In the embedding network design, AIGC personalized advertisement images and attributes are taken as inputs and put into the pre-trained VGG16 model to generate 128-dimensional feature vectors y. In order to make the AIGC personalized advertisement generation strategy more suitable for generation Z, the 128-dimensional vectors in the embedding network are used as the conditional information into the layout generative network (GAN), thus designing a VGG16-GAN-based AIGC personalized ad generation strategy model. Subsequently, on the basis of the neural graph cooperative model, the Adam algorithm is introduced, aiming to further improve the dissemination effect of its AIGC personalized advertisements, and in this way, the AIGC personalized advertisement recommendation graph neural network model (GNN-Adam) is constructed. The addition of VGG16 to the GAN model improves the PSNR and SSIM values of the generated AIGC personalized advertisement images, which range from 27.11 to 34.594 and 0.674 to 0.935, demonstrating the role of VGG16 in optimizing the generation strategy of AIGC personalized advertisements based on GAN. In addition, the MAE, Recall, and Precision results of AIGC personalized advertisement recommendation graph neural network model (GNN-Adam) are distributed as 0.011~0.048, 0.900~0.939, and 0.901~0.938, which reflects that Adam algorithm can enhance the dissemination effect of AIGC personalized advertisement. The research in this paper can provide valuable references for the theoretical research and design practice of AIGC personalized advertisement design under Generation Z, and prompt AIGC personalized advertisement to produce higher communication benefits.
